Karl Llewellyn on Legal Method: a Social Science Reconsideration

Tulsa Law Review

Volume 14 Issue 3

1979

Karl Llewellyn on Legal Method: A Social Science Reconsideration

R. H. Clark

Follow this and additional works at: https://digitalcommons.law.utulsa.edu/tlr

Part of the Law Commons

Recommended Citation R. H. Clark, Karl Llewellyn on Legal Method: A Social Science Reconsideration, 14 Tulsa L. J. 491 (2013).

Available at: https://digitalcommons.law.utulsa.edu/tlr/vol14/iss3/3

This Article is brought to you for free and open access by TU Law Digital Commons. It has been accepted for inclusion in Tulsa Law Review by an authorized editor of TU Law Digital Commons. For more information, please contact [email protected]. Clark: Karl Llewellyn on Legal Method: A Social Science Reconsideration

ESSAY

KARL LLEWELLYN ON LEGAL METHOD: A SOCIAL SCIENCE RECONSIDERATION

R.H. Clark*

I. INTRODUCTION One of the principal issues in any discussion of legal method is the amount of discretion enjoyed by trial and appellate judges. There has been, however, anything but unified agreement on the factors that lead judges to make particular decisions. The older mechanical, or "slot- machine," theory of judicial decision-making was put most effectively by Justice Roberts in 1936: When an act of Congress is appropriately challenged in the courts as not conforming to the constitutional mandate, the judicial branch of the Government has only one duty,-to lay the article of the Constitution which is invoked beside the statute which is challenged and to decide whether the latter squares with the former.' Essentially, Roberts argued that judges discover the law either in the timeless principles of the common law or in statutes and the Constitu- tion, and that a judge has no choice but to apply mechanically the su- perior norm. The rules of deductive logic guide the judge in applying the law once the correct premises are discovered. But during the same decades that the mechanical view of the judge's role flourished,' there were contrary and disturbing indications that this explanation was far too simple. For example, Cardozo pub-

* Trial Attorney, Antitrust Division, Justice Department, Washington, D.C. A.B., Univer- sity of California (Riverside), 1966; M.A., University of California (Santa Barbara), 1967, Ph.D., 1970; J.D., Capital University, 1977. The views expressed herein represent those of the author and not the Department of Justice. I. United States v. Butler, 297 U.S. 1, 62 (1936). See generally 0. ROBERTS, THE COURT AND THE CONSTITUTION (1951). 2. See note 11 infra and accompanying text.

Published by TU Law Digital Commons, 1978 1 Tulsa Law Review, Vol. 14 [1978], Iss. 3, Art. 3

TULSA LAW JO URIVAL [Vol. 14:491

lished several influential volumes during the 1920's which afforded a rare insight into the nature of judicial decision-making.3 His theory varied greatly with the view of Roberts. Cardozo was particularly effective in detailing the multiple pathways (logic, history, custom, and social welfare) by which an appellate judge could decide an important case. More significantly, beginning in the early 1930's, legal realism began to take hold in some leading American law schools.4 This theory provided a vital new way of looking at law and judicial policymaking. The realists, though divergent in orientation and methods, were unified in seeking to strip away the symbolism and self-serving declarations of judges and to focus instead upon the "real" factors leading judges to decide cases as they did. One of the principal leaders of this movement was Karl Llewellyn, who for thirty years was to investigate and to write about the actual functioning processes of judicial policymaking. Llew- ellyn dismissed the legal rules as "paper rules." Instead he maintained that such factors as judicial values and the evolution of law as it keeps pace with social change are more integral issues in understanding the legal process. In 1960, however, Llewellyn modified his original emphasis upon the freedom of action enjoyed by appellate judges. In his epic work, The Common Law Tradition: Deciding Appeals,5 Llewellyn laid out a number of "steadying factors" which he believed restrict the possible range of judicial discretion. Legal realism as a movement had sharply declined in importance by the early 1950's.6 A parallel investigation of the legal system by political scientists-the study of judicial behavior-proved to be more significant and continues to flourish.7 Using the insights and methodological tools of social science, the judicial behavior specialists took up and broadened the realists' investigation of the legal system. While they made greater use of statistical methods of analysis and of social science concepts, such as role theory and small group decision-making, their research substantiates the early proposals of the realists.' It has been suggested that the two movements are but separate stages in the

3. B. CARDOzo, THE NATURE OF THE JUDICIAL PROCESS (1921); B. CARDOZO, THE GROWTH OF THE LAW (1924). For a splendid collection of Cardozo's writings, see B. CARDOZO, SELECTED WRITINGS OF BENJAMIN NATHAN CARDOZO (M. Hall ed. 1947). 4. For a comprehensive discussion of the legal realists, complete with effective bibliography, see W. RUMBLE, JR., AMERICAN LEGAL REALISM (1968). 5. K. LLEWELLYN, THE COMMON LAW TRADITION: DECIDING APPEALS (1960) [hereinafter cited as LLEWELLYN, THE COMMON LAW TRADITION]. 6. See note 4 supra and accompanying text. 7. See, e.g., Ulmer, The Anavsis of Behavior Patterns on the United States Supreme Court, 22 J. POL. 629 (1960); Ulmer, Supreme Court Behavior and CivilRights, 13 W. POL. Q. 288 (1960). 8. See generally the authorities cited in note 45 infra.

https://digitalcommons.law.utulsa.edu/tlr/vol14/iss3/3 2 Clark: Karl Llewellyn on Legal Method: A Social Science Reconsideration

1979] K4RL LLEWELLYN same process of legal analysis, with the realists (and Llewellyn in particular) having blazed the trail which was later followed by the judicial behavioralists. 9 This paper seeks to reverse this process and to use the findings of the judicial behavior research to investigate and to substantiate or to reject some of the earlier realism theories, particularly the "steadying factors" suggested by Llewellyn. The result will be not only to evaluate the realist perspective as set out by Llewellyn, but also to demonstrate to the reader some of the insights pertaining to legal method that the two movements have uncovered.

II. LEGAL REALISM Any extensive discussion of the legal realists and their varied perspectives is beyond the scope of this essay.10 It is appropriate, however, to discuss briefly the general outlines of the legal realism perspective, and in particular to indicate Llewellyn's position within the movement. Legal realism was anticipated by several previous jurisprudential developments. First in importance was the early work of Oliver Wen- dell Holmes, Jr. Until challenged by Holmes, the primary American view of law had been drawn from Blackstone. According to Black- stone, law is a body of self-contained principles, true by virtue of their agreement with reason, and the answer to any legal problem rests in the correct logical deduction from this body of first principles.I' Law does not vary to match social evolution, and it is not really concerned with bringing about social change. The key goal is to maintain a logical nexus with the sacred body of principles that constitutes the common law. Statutory changes in or augmentations of the body of common law principles are to be resisted. Common law principles are timeless and can meet any emerging social problem, however novel. Holmes reacted against this position of "sterile logicism ' ' 2 in his essay, The Path of the Law, published in 1897.13 Similar views were expressed in his significant work, The Common Law:

9. Ingersoll, Karl Llewellyn, 4merican Legal Realism, and Contemporary Legal Behavioral- ism, 75 ETHICS 253 (1966). 10. Furthermore, the available literature on the movement is sufficiently abundant to obviate the need for any extended analysis. See note 4 supra. 11. For an excellent analysis of Blackstone's philosophy, see D. BOORSTIN, THE MYSTERIOUS SCIENCE OF THE LAW (1941). For American applications of Blackstone's thought, see P. MILLER, THE LIFE OF THE MIND IN AMERICA 156-71 (1965). 12. This term is drawn from Aronson, Tendencies in American Jurisprudence,4 U. TORONTO L.J. 90, 94 (1941). 13. Holmes, The Path ofthe Law, 10 HARV. L.J. 457 (1897).

Published by TU Law Digital Commons, 1978 3 Tulsa Law Review, Vol. 14 [1978], Iss. 3, Art. 3

TULSA LAW JOURNAL [Vol. 14:491

The life of the law has not been logic: it has been experience. The felt necessities of the time, the prevalent moral and political theories, intuitions of public policy, avowed or unconscious, even the prejudices which judges share with their fellowmen, have had a good deal more to do than the syllo- gism in determining the rules by which men should be gov- erned. The law embodies the story of a nation's development through many centuries, and it cannot be dealt with as if it contained only the axioms and corollaries of a book of mathe- matics. In order to know what is, we14 must know what it has been, and what it tends to become. Holmes believed that legal study should focus on the sources of legal rules, the historical antecedents of the rules, and changing social reali- ties which may require the alignment of legal rules and social con- duct. 5 Similar attacks on legal logic were later made in the writings of John Dewey, who saw logic not as "a deduction of certainties from theoretical or aprioriprinciples but a prediction of probabilities."' 6 A second vital precursor to legal realism was the sociological jurisprudence movement pioneered by Roscoe Pound. Pound's approach to law was founded on the pragmatic belief that good law satisfies social need. 7 He believed that the common law had become in part defec- tive, because it had stressed in an uncompromising fashion individual rights at the expense of social interests.' 8 For Pound, law is a tool for social engineering, 9 to assist in the realization of social need. Law is not something to be studied in books or from the standpoint of logical purity; rather, the proper subject of legal study should be the actual impact of law upon society.20 As a consequence, courts should not be limited by logical rules of deduction. Judges should take a broader perspective of social and economic circumstances, which would pro- vide a practical orientation based upon knowledge of economics, soci- ology, and politics. Law is a means to a greater end; law is not the end. The rules which comprise the common law are based on a past society

14. O.W. HOLMES, JR., THE COMMON LAW 5 (1963). 15. See Paul, FoundationsofAmerican Realism, 60 W. VIA. L.R. 37, 38-40 (1957). 16. Burrus, American LegalRealism, 8 How. L.J. 36, 39 (1962). See Dewey, Logical Method andLaw, 10 CORNELL L.Q. 17, 26-27 (1910). 17. E. BODENHEIMER, JURISPRUDENCE 110-12 (1974). On Pound, seegenerall, D. WIGDOR, ROSCOE POUND: PHILOSOPHER OF LAW (1974). 18. This theme is probably best demonstrated in R. POUND, THE SPIRIT OF THE COMMON LAW (1963). 19. W. FRIEDMANN, LEGAL THEORY 336-37 (5th ed. 1967). 20. Savarese, American Legal Realism, 3 Hous. L.R. 180, 184 (1965).

https://digitalcommons.law.utulsa.edu/tlr/vol14/iss3/3 4 Clark: Karl Llewellyn on Legal Method: A Social Science Reconsideration

1979] KARL LLEWELLYN

with different needs. New conceptions of property, contract, and individual rights are necessary, and the law has to adjust to these new demands. A considerable amount of overhauling of legal theory was in order, for only by bringing law into balance with social needs could law serve its proper functions. Against this background, the legal realism movement began to develop in the 1930's.1 It is certainly incorrect to suggest that there was a single legal realism movement. In actuality, the movement was made up of individuals who shared some common viewpoints toward law, but who maintained highly individualistic perspectives as well.22 There was no set orthodoxy around which the realists tended to congregate, other than a desire to study empirically the actual operation of the law. Two principal branches of the movement emerged, one led by Jerome Frank and the other by Karl Llewellyn. In fact, Frank recognized this division in the 1948 preface to his monumental Law and the Modern Mind23 in which he classified Llewellyn's perspective as that of the 24 "rule skeptic." The rule skeptics branch of realism focused upon the appellate courts and the role that legal rules play in limiting their discretion in decision-making. Drawing from Pound's emphasis upon law in action, as well as from the attacks of Holmes and Dewey upon legal logic, the rule skeptics were doubtful that case law precedent had any restraining effect on the judges. They pointed to such factors as the multiplicity of precedents, the various accepted techniques for manipu- lating precedents, the inability of logic to indicate which of a number of competing premises ought to be selected, the ambiguity of legal language, and the role of psychological influences (which one realist judge termed "hunches"25) in making decisions.26 Llewellyn suggested that in reality legal rules were frequently nothing more than "paper" rules which often suggested what ought to be, rather than being the actual result of judicial decisions.27 Facts, as manifested in what judges actually did, were the vital focus of legal study, and not abstract rules which supposedly dictated a particular result. As a consequence, law was an

21. See note 4 supra. 22. K. LLEWELLYN, JURISPRUDENCE 53 (1962) [hereinafter cited as LLEWELLYN, JURISPRU- DENCE]. Llewellyn recognizes some of the common perspectives shared by the bulk of the realists to be discussed in this article. Id. at 55-60. 23. J. FRANK, LAW AND THE MODERN MIND (1963). 24. Id. at xi. 25. Hutcheson, The Judgment Intuitive: The Function of the "Hunch"in JudicialDecision, 14 CORNELL L.Q. 274 (1929). 26. W. RUMBLE, JR., supra note 4, at 50-78. 27. See LLEWELLYN, JURISPRUDENCE, supra note 22, at 22-25.

Published by TU Law Digital Commons, 1978 5 Tulsa Law Review, Vol. 14 [1978], Iss. 3, Art. 3

TULSA LAW JOURNAL [Vol. 14:491

empirical rather than a normative science.2" In short, wrote Llewellyn, "there is less possibility of accurate prediction of what courts will do than the traditional rules would lead us to suppose."2 9 Therefore, the rule skeptics-while certainly not dismissing the role of rules in all situations-felt that judges are free to choose between conflicting prece- dential premises. Judicial decisions, no matter how fully clothed in logic and abstract rules, do not candidly explain why one particular premise had been chosen over others. As a result, law is what officials do, not what they say, and this should be the orientation of legal study. The second faction of the realist movement will be mentioned only briefly here, because its full development is tangential to this article. The "fact skeptics," led by Frank, focused upon trial rather than appellate courts. Their criticisms were far more devastating than simply sug- gesting that legal rules did not always dictate the outcome of a case. Frank's thesis was that the courts cannot even make accurate findings of fact, let alone apply the correct rules to those facts.30 He believed that "the certainty of law is an illusion, a fictitious, unconscious pre- tense to satisfy the infantile craving of the public for security."'31 Frank argued that the frailties of human memory and language, coupled with the subconscious desires of witnesses, jurors, and judges, combine to produce a situation where facts are seldom accurately ascertained by the trial process. These problems were further accentuated by the adversary or "fight theory" of trial procedure. When the goal is to win, relevant facts may be kept out of the trial completely. 2 Hence, even if rules are the motivating force in judicial decision-making, there still is little possibility for having a truly rule-oriented system of law. Further- more, the judges' own subconscious psychological desires and intuitions are far more influential in making decisions than is any body of 33 rules. The rule skeptics were met with a flood of strong criticism in response to their theories of judicial decision-making. Expected adverse comment emanated from those who felt that legal rules and logic are

28. See Moskowitz, The American Legal Realists andan Empirical Science of Law, I1 VILL. L.R. 480, 498-501 (1966). 29. LLEWELLYN, JURISPRUDENCE, supra note 22, at 60. 30. See generally J. ROSENBERG, JEROME FRANK: JURIST AND PHILOSOPHER (1970). 31. Yntema, American Legal Realism in Retrospect, 14 VAND. L.R. 317, 321 (1960). 32. See generallyJ. FRANK, COURTS ON TRIAL: MYTH AND REALITY IN AMERICAN JUSTICE (1949). 33. Purcell, American Jurisprudencebetween the Wars: Legal Realism andthe Crisis of Deno- cratic Theory, 75 AM. HIST. REV. 424, 431-32 (1969).

https://digitalcommons.law.utulsa.edu/tlr/vol14/iss3/3 6 Clark: Karl Llewellyn on Legal Method: A Social Science Reconsideration

1979] K4RL LLEWELLYN

far more influential in decision-making, if not determinative, than the realists were willing to admit.3" Another major source of criticism, however, was somewhat surprising. Roscoe Pound published an influential attack upon the realists in 1931.11 He criticized the realists for their extreme underemphasis of the role of rules, 36 and for their reli- 37 ance upon statistical documentation of what the courts actually do. Pound's most basic point was that the realists, by concentrating exclusively upon what the courts do in fact, have ignored the vital issue of what the courts ought to do.38 Pound and the sociological jurisprudence movement had been deeply concerned with realizing social justice through the use of law. The rule skeptics appeared to Pound to throw over all normative issues and to focus only upon the empirical analysis of what judges in fact do. Llewellyn responded by asserting that this division of the "is" from the "ought" was only tempo- rary-that to make any reforms in the legal system it is necessary to demonstrate empirically what the system is actually doing.39 A third criticism, levied by Lon Fuller, attacked the realists for their easy rejection of legal rules and fervent adoption of behavioralism.4° For Fuller, realism's rejection of universals and abstractions tended to discount the role that concepts play in legal reasoning. Fuller was supported by Kantorowicz, who argued that a certain degree of discretion is inherent in any legal rule or concept and that the realists, by over-emphasizing the amount of that discretion, had merely set up a

34. See, e.g., M. COHEN, LAW AND THE SOCIAL ORDER 165 (1967); J. DICKINSON, ADMINIS- TRATIVE JUSTICE AND THE SUPREMACY OF LAW IN THE UNITED STATES 128-40 (1959); Cohen, The Place ofLogic in the Law, 29 HARV. L.R. 622 (1915); Dickinson, Legal Rules: Their Function in the Processof Decision, 79 U. PA. L.R. 833 (1931). One could also, of course, point to the "pure theory of law" as manifesting the viewpoint that legal rules alone can serve as an effective device with which to study the legal system. See Kelsen, The Pure Theory ofLaw (pt. 1), 50 LAW Q. RE. 474 (1934) and (pt. 2) 53 LAW. Q. REV. 517 (1935). See also W. EBENSTEIN, THE PURE THEORY OF LAW (1945). For an effective position asserting the opposing viewpoint, see Stoijar, The Logi- cal Status ofa LegalPrincple,20 U. CHI. L.R. 181 (1953). It was recently asserted that some new developments in analytical positivism have led to a rapprochement with legal realism. See Bodenheimer, Analytical Positivism, Legal Realism, and the Future of Legal Method, 44 VA. L.R. 365 (1958). 35. Pound, The Call/ora Realist Jurisprudence, 44 HARV. L.R. 697 (1931). 36. Id. at 705-07. 37. Id. at 701-02. 38. Id. at 700. 39. LLEWELLYN, JURISPRUDENCE, supra note 22, at 55-56. For further rebuttal of Pound's position, see Jones, Law and Morality in the Perspective ofLegal Realism, 61 COLUM. L.R. 799 (1961). 40. Fuller, American Legal Realism, 82 U. PA. L.R. 429, 445-47 (1934). Fuller's attack is by no means limited to this point. See L. FULLER, THE LAW IN QUEST OF ITSELF 52-64 (1966).

Published by TU Law Digital Commons, 1978 7 Tulsa Law Review, Vol. 14 [1978], Iss. 3, Art. 3

TULSA LAW JOURVAL [Vol. 14:491

straw man to knock down.4 ' Irrespective of these criticisms, the realists did have an impact upon legal thought. Grant Gilmore has suggested that the legal activity which has produced the Restatements, the Ameri- can Law Institute, and various movements toward codification, indicates the significance of the realist perspective.4 z That legal realism had a significant impact on legal theory is beyond question. By the middle 1950's, Llewellyn had reached the conclusion that the public in general, and the bar in particular, had become so skeptical of judicial decision-making that they had ignored totally the restraining role that rules do play. A genuine "crisis of con- fidence" faced the appellate courts, and Llewellyn decided to write a book explaining the elements which he felt limit appellate judicial discretion. This significant volume, The Common Law Tradition:Deciding Appeals, published in 1960, was devoted to a study of these restraining influences. Llewellyn identified fourteen "steadying factors" which he saw as limiting the free scope of judicial discretion.43 In addition, he supplemented his thesis by extensive case studies of different appellate courts' policymaking periods. While the book does not stress social science methods, and while a social scientist might question the accuracy of the sampling technique employed by Llewellyn, he does refer to concepts familiar to today's social scientists. This paper, after an introduction to the orientation and techniques of judicial behavior studies, will discuss the validity of these asserted "steadying factors" as they have been subsequently investigated in the political science literature. While not all of Llewellyn's theses will be investigated, the paper will focus upon four or five of his most suggestive and significant theses. It should also be emphasized that Llewellyn did not abandon his realist perspective, but he sought to present a more complete picture of the factors influencing judicial decision-making by discussing the restraining factors that he believed play an important role in that decision-making. 4

III. THE JUDICIAL BEHAVIOR MOVEMENT The focus of political science has shifted from institutional to indi-

41. Kantorowicz, Some RationalismAbout Realism, 43 YALE L.J. 1240, 1244-45 (1934). On this issue of the role of discretion in law, see generally K. DAVIS, DISCRETIONARY JUSTICE (1969). 42. Gilmore, Legal Realism Its Cause and Cure, 70 YALE L.J. 1037 (1961). See also G. GILMORE, THE DEATH OF CONTRACT 58-59 (1974). 43. LLEWELLYN, THE COMMON LAW TRADITION, supra note 5, at 19. 44. See L. LLOYD OF HAMPSTEAD, INTRODUCTION TO JURISPRUDENCE 409-13 (3d ed. 1972).

https://digitalcommons.law.utulsa.edu/tlr/vol14/iss3/3 8 Clark: Karl Llewellyn on Legal Method: A Social Science Reconsideration

19791 KARL LLEWELLYN

vidual behavior.45 Paralleling this movement, attention has been given, within the public law sub-category, to the judicial process-the methods of judicial decision-making-rather than simply to the analysis of the product of that process, case decisions. Frankly recognizing that judges are makers of public policy, recent research has taken tentative steps toward the creation of a framework of analysis that explains both what decisions have been made and how such decisions result from a judicial policy process. Research studies have examined a wide range of relevant variables and have related them to the final decisions. Such factors as where judges come from, how their backgrounds influence their judicial behavior, the role of the judicial environment in affecting behavior, strategies of intra- and extra-court influence, and the relationship of the courts to other policymaking institutions have all come under examination.46 More precisely, recent judicial process research has attempted to achieve certain basic goals. Theodore L. Becker has concisely defined these objectives as (1) precise, quantitative descr~ition; (2) social scien- tific explanation of political behavior; and (3) reasonably precise, quan- titativepredictionsof political behavior.47 Or, as Glendon Schubert has suggested, "Judicial behavioralism is an attempt to construct a system- atic theory about human behavior, analyzing data about judges and adjudicatory processes of decision-making by using theories and methods from all the behavioral sciences, according to their relevance to the particular inquiry at hand. '48 Of the goals indicated, explanation is the most cherished and valuable. Unfortunately, it is also the most difficult to attain. The scholarly attempt to relate certain background characteristics of judges to their decision-making patterns has been one of the first and most widely employed methods of studying the judicial process. The focusing upon factors other than logical deduction immediately indicates the close kinship of the study of judicial behavior with legal real-

45. The judicial behavior movement is traced in Pritchett, Public Law and JudicialBehavior, 30 J. POL. 480 (1968); Rigby & Witt, BibliographicEssay. BehavioralResearch in PublicLaw 1963- 1967, 22 W. POL. Q. 622 (1969); Schubert, BibliographicEssay: Behavioral Research in Public Law, 57 AM. POL. Sci. REv.433 (1963); Shapiro, PoliticalJurisprudence, 52 KENTUCKY L.J. 294 (1964). 46. See generally authorities cited in note 45 supra. 47. Becker, Inquiryinto a School of Thought in the JudicialBehavior Movement, 7 MIDWEST J. POL. Sc. 254, 256 (1963). The research techniques developed to achieve these goals include bloc analysis, role theory, jurimetrics, small-group theory, attitudinal scaling, systems analysis, and social background analysis. These will be examined more fully below. 48. Schubert, Ideologies and Attitudes, Academic and Judicial,29 J. POL. 3, 7 (1967).

Published by TU Law Digital Commons, 1978 9 Tulsa Law Review, Vol. 14 [1978], Iss. 3, Art. 3

TULSA LAW JOURNAL [Vol. 14:491

ism. One early attempt to use this method was Charles Groves Haines' pioneering study.4 9 In seeking to strike a reasonable mean between ad- vocates of "slot-machine" jurisprudence and those favoring the "free legal decision" viewpoint, Haines concisely stated what this type of study seeks to establish: Just as is the case with other opinions of individuals, judicial opinions necessarily represent in a measure the personal impulses of the judge, in relation to the situation before him, and these impulses are determined by the judge's lifelong series of previous experiences." Basically, background studies attempt to identify and then to relate such factors as personality, education, family and personal associations, political affiliations, and judicial temperament to judges and their patterns of decision-making. Unfortunately, a warning provided by Haines is often overlooked: "The influences which move men even to do simple acts are too subtle to be summed up in such a catchword as 'economic bias.' "151 Haines was content to cite a few 1916 statistics on the sentencing behavior of New York magistrates to indicate that variation can result because judging is "a personal thing." More recent studies have utilized sophisticated statistical devices. Two articles by Stuart Nagel demonstrate this technique. In the first,52 Nagel was able to establish a statistically significant relationship between party membership and judicial votes in criminal cases during 1955 on the New Jersey and Penn- sylvania Supreme Courts. Briefly, Nagel demonstrated that Democrats were above their courts' averages in deciding in favor of defendants on appeal, while Republicans were below their courts' averages. In the second article,53 similar techniques were used to relate party membership to fifteen types of legal issues in a sample of 298 state and federal supreme court judges. Once again, in nine categories Nagel demonstrated a statistically significant relationship between party membership and voting. The Democrats were generally above their courts' averages in proliberal decisions (ie., favoring lower classes and social change), while the Republicans were below the average. Nagel is cautious in his

49. Haines, GeneralObservations ofthe Effects of Personal,Political and Economic Influences in the Decision of Judges, 17 ILL. L.R. 96 (1922). 50. Id. at 104-05. 51. Id. at 116. 52. Nagel, Testing Relations Between Judicial Characteristicsand Judicial Decision.Making, 15 W. POL. Q. 425 (1962). 53. Nagel, Political Party Affliation andJudge's Decisions,55 AM. POL. Sc. REV. 843 (1961).

https://digitalcommons.law.utulsa.edu/tlr/vol14/iss3/3 10 Clark: Karl Llewellyn on Legal Method: A Social Science Reconsideration

19791 K4RL LLEWELLYN interpretation of these findings, saying only that party affiliation and the nature of decisions both stem from the same background factors. But he does indicate a well-founded belief that such a technique can be used to test broad propositions relating to the interaction between background factors and judicial decisions. Similar fundamental techniques have been used by John Schmidhauser. In a study54 covering the Supreme Court from 1790 through 1957, Schmidhauser related adherence to precedent, overruling of precedent, and dissents from majority opinions, to such factors as party affiliation, regional background, and prior judicial experience. Using case evaluation techniques similar to those used by Nagel, Schmidhauser demonstrated that certain combinations of background factors were present for upholders of precedent, for overrulers, and for dissenters. In a second study,55 covering the Supreme Court from 1837 through 1860, Schmidhauser suggested that party affiliation, rather than regional background, was a factor in judicial behavior on the Ta- ney Court in relation to slavery and business decisions. To prove that two facts are present in certain situations is not to prove that one factor causes the other. Indeed, the cautious caveats of Nagel and Schmidhauser emphasize this point. Joel Grossman eyed this technique critically,5 6 and his observations are both relevant and interesting. One of his basic concerns is that these studies tacitly as- sume that factor A causes result B, without being concerned with the possible influences of other factors in this "conversion" process. Sec- ond, such studies ignore judicial consensus by gathering as raw data only nonunanimous opinions. Since most judges on most courts agree far more than they disagree, institutional influence is too significant to be disregarded. Third, Grossman doubts whether clearcut value choices are frequently available to judges. He believes that it is more likely that they have a series of marginal choices. Because each background factor is presumed to have equal effect on each judge in each judicial decision and because each judge is studied in isolation, the real-life situation in which judicial decision-making takes place can be distorted.

54. Schmidhauser, Stare Decisis, Dissent, and the Background ofthe Justices of the Supreme Court ofthe United States, 14 U. TORONTO L.R. 194 (1962). 55. Schmidhauser, JudicialBehavior and the Sectional Crisis, 1837-1860, in CONSTITUTIONAL LAW IN THE POLrricAL PROCESS 486 (1963). 56. Grossman, Social Backgrounds andJudicial Decisions: Notesfor a Theory, 29 J. POL. 334 (1967); Grossman, Social Backgrounds andJudicial Decision-Making, 79 HARv. L.R. 1551 (1966).

Published by TU Law Digital Commons, 1978 11 Tulsa Law Review, Vol. 14 [1978], Iss. 3, Art. 3

TULSA LAW JOURNAL [Vol. 14:491

IV. LLEWELLYN'S STEADYING FACTORS A. ProfessionalJudicial Office The most important steadying factor, Llewellyn suggests, is what he terms "professional judicial office."57 By this he means that the individual who occupies a judicial office will have a certain normative model against which to measure his own behavior. Such factors as im- partiality, uprightness, fairness, and susceptibility to reasoned argument are components of the judge's role. As a result, the individual judge seeks to comply with a model of how a judge ought to decide, even if this involves a diminishing of what his own values tell him he should decide. As Llewellyn puts it, "The office waits and then moves with majestic power to shape the man." 8 Therefore, role theory can explain how judges are in fact restrained by their own conception of judicial office, a far more significant restraint than precedent or logic. In the wake of the important work being done in the legislative process involving the concept of role theory, it was probably inevitable that such a research technique would be utilized to study the judicial process. Kenneth Vines has defined role as a "description of the normative expectations concerning the proper behavior and personal qual- ities of a specific social or political position."' 9 The normative aspect is vital to the concept of role;60 behavior which is in accord with such normative expectations is designated as "role behavior." The role concept is valuable in that it allows evaluation in terms of the institutional framework within which individual behavior takes place. Vines has made one tentative exploration into the feasibility of using role theory in judicial research.61 He suggests that role theory may be particularly valuable in judicial research since there are many possible roles which a judge can legitimately play within the formal and traditional standards of the bench. With this in mind, Vines studied four state supreme courts for one year. The first part of Vine's study drew its data from questionnaires submitted to the judges. From this sample, Vines posited a number of judicial models of role behavior. For example, some judges saw their role as one of settling disputes,

57. LLEWELLYN, THE COMMON LAW TRADITION, supra note 5, at 45. 58. Id. at 48. 59. Vines, The JudicialRole in the American States, in FRONTIERS OF JUDICIAL RESEARCH 464 (J. Grossman & J. Tanenhaus ed. 1969) [hereinafter cited as Vines]. 60. J. WAHLKE, H. EULAU, W. BUCHANAN, & R. FERGUSON, THE LEGISLATIVE SYSTEM: EXPLORATIONS IN LEGISLATIVE BEHAVIOR 9 (1962). 61. Vines, supra note 59.

https://digitalcommons.law.utulsa.edu/tlr/vol14/iss3/3 12 Clark: Karl Llewellyn on Legal Method: A Social Science Reconsideration

1979] KARL LLE WELL YN

others suggested one of supervising the lower courts, while still others envisioned themselves as defenders of the Constitution. Some judges frankly admitted that they made policy and felt it was necessary to dis- charging their responsibilities, while others saw their legitimate function as being narrowly limited to strict interpretation of the law. In the second part of the study, Vines utilized case analysis to check the agreement between the role that the judge believes he fulfills and his actual behavior. Vines used unanimous opinions in this study, as well as nonunanimous decisions, and found that often members with the same party identification perceived their roles significantly differ- ently. His results suggest that the background analysis technique, in confining its data to nonunanimous opinions only,62 has exaggerated the degree of intracourt disagreement and has been consequently mis- led in its findings. Vines concludes that his study has far too little data for any definitive findings, his purpose being merely to demonstrate the potential utility of role analysis. A second example of role theory is found in Joel Grossman's examination of Justice Frankfurter.63 This work is a reaction to various studies that were based upon vote counting and that indicated Frank- furter was unsympathetic to civil liberty claims. Grossman felt that such a uni-dimensional analysis "ignores this complex of decision- making factors and concentrates almost exclusively on the result, which is only a part of the entire decision-making process."' Grossman believed that role theory can give a more accurate and comprehensive picture when added to these other findings. To ascertain Frankfurter's perception of his own role, Grossman extracted propositions from his opinions which showed Frankfurter's tendency toward judicial restraint.65 Grossman concluded that, while Frankfurter undoubtedly voted on occasion against civil liberty claims based upon his perception of judicial restraint rather than substantive opposition to such claim- ants, extraction of those cases shows Frankfurter to be a few steps closer to the center on such issues. Grossman's point is well made.

62. See note 56 supra and accompanying text. 63. Grossman, Role-Playing and the Analysis of JudicialBehavior: The Case ofMr.Justice Frankfurter, II J. PUB. LAW 285 (1962). 64. Id.at 286. 65. Grossman defines role as "a consistent pattern of behavior on the part of an individual in response to his conception of the nature of his function in a system." Id. at 294. For other conceptions of role, see ROLE THEORY (B. Bidde & E. Thomas eds. 1966); N. GRoss, W. MASON, & A. MCEACHERN, EXPLORATIONS IN ROLE ANALYSIS: STUDIES OF THE SCHOOL SUPERINTENDENCY ROLE (1958).

Published by TU Law Digital Commons, 1978 13 Tulsa Law Review, Vol. 14 [1978], Iss. 3, Art. 3

TULSA LAW JOURNAL [Vol. 14:491

Uni-dimensional scaling (based on vote counting) can be dangerous if confined to a single-factor type analysis; role theory can offer a valuable tool to complement such techniques. A Vines-type questionnaire methodology was used by Theodore Becker in his study of thirty Hawaiian judges. 66 He determined that certain factors (such as the judge's perception of justice in the case; the appearance of highly respected counsel; or the judge's belief of what the public welfare required) influence the role of judicial decision- maker. Becker was also concerned with the role adherence to precedent played in influencing judicial decisions. In the first part of the questionnaire, he asked the judges to rank, in order of importance, factors that influenced their decisions in a case. Next, Becker posited a hypothetical case and asked for a decision. Thus, like Vine's study of case decisions,67 Becker sought to compare judicial perception with actual behavior. His findings indicated that 68% rated subjective factors such as "common sense" and "what the public demands" as more influential than adherence to clear precedent. Yet, in deciding the hypothetical case, 78% held precedent to be the single most important factor. Again, this study is merely a demonstration by Becker of techniques, and it used only a small sample of data for illustration. The most extensive use of role theory to explain judicial behavior is found in Henry Glick's Supreme Courts in State Politics.68 Drawing his data from a number of state supreme courts, Glick was able to identify several different roles that judges saw themselves as playing: lawmaker (policymaker), law interpreter, protector of individual rights, and administrator of state court systems. The conflict between the role of lawmaker and that of law interpreter is most important. Some judges openly expressed their opinion that it was their responsibility to make policy when the occasion arose. Glick suggests that this view is more likely in a state manifesting a liberal political orientation, such as Pennsylvania. In contrast, the law interpreters were more passive and devoted to implementing narrowly either past precedent or statutory legislation. A conservative political environment, such as Louisiana, encouraged this role orientation. The law interpreters rated precedent as a much more vital factor than did the lawmakers, who instead ranked social welfare and justice as more vital influences.

66. Becker, A Survey Study ofHawaiianJudges.- The Effect on Decisions ofJudicialRole Par!. aions, 60 Am. POL. Sci. REV. 677 (1966). 67. See note 61 supra and accompanying text. 68. H. GLICK, SUPREME COURTS IN STATE POLITICS (1974).

https://digitalcommons.law.utulsa.edu/tlr/vol14/iss3/3 14 Clark: Karl Llewellyn on Legal Method: A Social Science Reconsideration

1979] KRL LLEWELLYN

This recent judicial research affirms Llewellyn's thesis that judicial role does act as a factor of influence in decision-making. There is even some indication that role concepts carry over into nonjudicial activities of judges as well.6 9 Some work has been done using role theory as a tool of analysis for studying trial judges also.7° It is imperative to note, however, that role is not always an influence which restrains judges from innovative decision-making. For the lawmakers, role concept would encourage the flexibility and discretion that Llewellyn felt had been overstressed by critics of appellate courts. It is clear, however, that appellate decision-making is not simply a stimulus (value preference)-response (decision) relationship. Another variable, the judge's own conception of how he ought to decide, intervenes and exercises some vital influence.

B. Collegial Decision-Making A second major steadying factor cited by Llewellyn is "group decision.' Critics often overlook the fact that judges at the appellate level make their decisions in groups, either in panels of three or groups of nine at the intermediate appellate level, and by full bench at the Supreme Court level. As a result, Llewellyn argued, a judge is restrained from exercising absolute discretion by having to convince other judges of the merits of his position. In addition, in writing opinions, judges write for the others in their majority or dissenting bloc.7" Therefore, an opinion is not simply the rationalization of an individual judge, but it is rather a joint product. Cooperative brief-writing also diminishes the discretion available to any single appellate judge. Having utilized background influence analysis and posited various types of perceived judicial roles, some students of the judicial process have turned to the sociological tool of small group theory. As defined by Eloise C. Synder, small group theory "deals with the Supreme Court as a small group and attempts to discover what group processes are present as the justices solve the important problems brought before them.' 73 It is this process of face-to-face interaction, with the possible

69. Beiser, Goodman, & Cornwell, JudicialRolein a NonjudicialSetting, 5 L. & Soc'Y REV. 571 (1971). 70. Ungs & Baas, JudicialRole Perceptions: .4 Q-Technique Study of Ohio Judges, 6 L. & Soc'Y REv. 343 (1972). 71. LLEWELLYN, THE COMMON LAW TRADITION, supra note 5, at 31. 72. Id. at 26. Though Llewellyn tends to identify opinion-writing as a separate factor, the author believes that it rightly should be considered as an aspect of collegial decision-making. 73. Synder, The Supreme Court as a Small Group, 36 Soc. FORCES 232 (1958).

Published by TU Law Digital Commons, 1978 15 Tulsa Law Review, Vol. 14 [1978], Iss. 3, Art. 3

TULSA LAW JOURNAL [Vol. 14:491

manifestation of cliques, collegial leadership, patterns of influence, bargaining, and tactics of persuasion, that is the focus of this method.74 The existence of persistent patterns of similar voting behavior among Court members is at least a tentative indication that small group, interpersonal processes are at work. A pioneering study using this technique was done by Eloise Synder in 1958." 5 The goal of her work, which covered the Supreme Court during the period of 1921-1953, was to determine whether sub- groups or cliques were evident from analysis of case opinions. Synder reported patterns of agreement and disagreement among Court members over a long period, although agreement was much more frequent. She found three blocs: a liberal (e.g., 1943-1945: Black, Douglas, Rut- ledge, Murphy), a conservative (e.g., 1930-1932: Butler, Van Devanter, Sutherland, McReynolds), and a moderate-pivotal (e.g., 1943-1945: Stone, Reed). Although the membership changed due to attrition, these blocs persisted. Some change did take place, but with the excep- tions Frankfurter and Jackson that modification was limited to shifts from either the liberal or the conservative to the pivotal bloc. These moves were seldom completely to the opposite pole on the continuum. Synder believed that these shifts, mostly from liberal to pivotal, were a result of changing "conceptual frameworks" due to new Court person- nel. When newer radical ideas came to the Court, the older liberals, holding to their tenents, were in effect moved toward the conservative pole. The late Justice Black may have been a current example of this phenomenon. Small group analysis must be accepted carefully. As Walter Mur- phy has indicated: Since Pritchett and Synder only used voting records they could discover little more than that Justices could be classified; study of groups also requires consideration of interpersonal interaction and influence. The fact that two or more Justices vote together is rather weak evidence that their votes are the result of interaction; standing alone, voting records tell very little about the force or direction of any interpersonal influence that may exist. Small group analysis requires other

74. See C. PRITCHETr, THE ROOSEVELT COURT (1969); C. PRITCHETT, CIVIL LIBERTIES AND THE VINSON COURT (1954). This particular methodological development grew, in part, out of the early efforts of Pritchett and others to demonstrate the persistence of blocs on the Court through vote analysis. 75. Synder, supra note 73.

https://digitalcommons.law.utulsa.edu/tlr/vol14/iss3/3 16 Clark: Karl Llewellyn on Legal Method: A Social Science Reconsideration

1979] 9KRL LLEWELLYN kinds of data and a more general understanding of the impact of a group decision situation on individual behavior.76 Since the judiciary is fairly well removed from direct scholarly observa- tion, such innermost penetration presents great difficulties. Neverthe- less, several more sophisticated studies-based upon papers of Justices, law clerks' observations, and perhaps even some rumors-have been forthcoming. One prime example is provided by David Danelski, who evaluated group leadership from the perspective of the task leader and of the social leader in his study of the Chief Justices during conference.77 A task leader leads the group to make decisions; the social leader works to maintain social cohesion within the group. Drawing his data primarily from papers of the Justices, Danelski posits four basic situations that can develop. A Chief Justice can be both social and task leader (Hughes), social but not task leader (Taft), task but not social leader (no example given), or neither social nor task leader (Stone).78 The ideal situation is the firm leadership of Hughes, which produced low intra-group conflict, high social cohesion, satisfaction with conference, and high productivity. The worst situation resulted from lack of any leadership (Stone) which produced unrestrained bickering and reduced cohesion, satisfaction, and productivity. The Chief Justice may also influence his group through opinion assignment and maximization of group unity. In assigning opinions, Danelski suggests two possible strategies which the Chief Justice may follow: Rule 1: Assign the case to the justice whose views are the closest to the dissenters on the ground that his opinion would take a middle approach upon which both majority and minority could agree. Rule 2: Where there are blocs on the Court and a bloc splits, assign the opinion to a majority member of the dissenters' bloc on the grounds that (a) he would take a middle approach upon which both majority and minority could agree and (b) the minority justices would be more likely to agree with him because of general mutuality of agreement.79

76. Murphy, Courts as Small Groups, 79 HARv. L.R. 1565-67 (1966). 77. DANELSKI, The Influence of the Chief ustice in the DecisionalProcess, in C. PRITCHETT & D. DANELSKI, COURTS AND JUDGES 497-508 (1961). 78. For a full discussion of these firm leadership techniques, see McElwain, The Business of the Supreme Court as Conducted by Chief Justice Hughes, 63 HARV. L.R. 5-26 (1949). 79. Danelski, supra note 77, at 504.

Published by TU Law Digital Commons, 1978 17 Tulsa Law Review, Vol. 14 [1978], Iss. 3, Art. 3

TULSA LAW JOURN4L [Vol. 14:491

In uniting the Court, the Chief Justice tries to limit dissent and to pro- mote intra-group compromise. This spirit of moderation and compromise eases tense arguments and allows overall agreement and productivity to rise. Moving the focus away from the Chief Justice, Walter Murphy, again working from private papers, has contributed a most fascinating article on how Justices may seek to influence each other through the use of social skills.10 The study indicates some of the methods by which a Justice can mass support behind his position on controversial issues. Murphy cautions the reader, however, that the Court is a particular type of group: it works within a well-defined and peculiar tradition, and its members must look forward to a long period of future cooperation. The Chief Justice is expected to lead, and he can consequently be more direct than the associate Justices. However, all Jus- tices can influence their colleagues by rational persuasion, by cordial welcome for all new members, by favorable comments on others' work, by common courtesy, and by compromise and mutual accommodation on occasion. The most precious instrument of influence is a Justice's vote and threat of dissent. Bargaining, done tenderly as not to antago- nize, becomes a crucial social skill. Murphy concludes that a socially skillful Justice cannot create a bloc but that adept use of social skills can be utilized to reinforce ideological affinities that already exist. Sidney Ulmer has contributed some valuable insights into the role of precedent in judicial decisions in his study of the Supreme Court's history of overruling its own precedents. 8 Looking at the Court's history through fairly traditional research devices, Ulmer explored a number of propositions relating to the Supreme Court's overruling itself. He found that the Court has only overruled specifically-in contrast to distinguishing or ignoring precedent-in cases where a majority agreed. Commerce and taxation precedents have been most often overruled, and overruling has usually been done to emphasize a change rather than as a final step in a long evolution. Mathematically the cases overruled had neither any correlation with the number of opportunities to overrule, nor with the party affiliations of Justices. Apparently, the most crucial factor was the appointment of many new members to the Court within a short time. Thus, the changing composition of the

80. Murphy, Marshalingthe Court: Leadershp, Bargaining,and the Judicial Process,29 U. Cm. L.R. 640 (1962). 81. Ulmer, An EmpiricalAnalysis of Selected Aspects of Lawmaking of the United States Supreme Court, 8 J. PuB. L. 414 (1959).

https://digitalcommons.law.utulsa.edu/tlr/vol14/iss3/3 18 Clark: Karl Llewellyn on Legal Method: A Social Science Reconsideration

1979] K4ARL LLEWELLYN

judge's decision-making group may well influence his own perceptions of adhering to or deviating from precedent. Some work on the federal courts of appeal by Sheldon Goldman indicated that these techniques of analysis can be applied to smaller appellate decision-making groups as well.82 Goldman studied the behavior of the eleven circuit courts of appeal in search of some fundamental generalizations. He found that liberal and conservative blocs generally do exist on these lower appellate benches as well. Party affiliation was found to be most closely related to voting behavior, espe- cially in cases involving economic liberalism. Other variables, such as social-economic origin and standing, education, and age, were not found to be significantly related to voting behavior. While the article's scope is not broad, it does serve to initiate some understanding of the role which collegial decision-making plays in the intermediate federal courts of appeal. Llewellyn rested his argument that collegiate activity affects judicial decision-making upon a point subsequently well explored in the political science literature. As Charles H. Sheldon has indicated, "the basic assumption underlying micro-group models is that when men (judges) get together into some structured situation (collegial court), their behavior is altered."83 This literature demonstrates the validity of Llewellyn's thesis that collegial decision-making can alter what the sol- itary judge might decide. However, this still is no guarantee that in any given situation that bloc will not decide to play a strong policymaker rather than law interpreter role. Yet clearly Llewellyn is correct in pointing to the influence of consensual decision-making as a restraint upon individual judicial values.

C. A Known Bench A third important factor suggested by Llewellyn is that of a known bench. 84 By this term Llewellyn meant that over a period of time the performance of an appellate court-and its constituent members--can be studied by lawyers and that they can discover the ways a particular bench has looked at issues, how it has handled precedents, and most importantly its attitudes. As Llewellyn put it, opinions "cumulate to

82. Goldman, Voting Behavior on the United States Courts ofAppeals, 1961-1964, 60 AM. POL. Scr. REV. 374 (1966). 83. C. SHELDON, THE AMERICAN JUDICIAL PROCESS: MODELS & APPROACHES 51 (1974). 84. LLEWELLYN, THE COMMON LAW TRADITION, supra note 5, at 34.

Published by TU Law Digital Commons, 1978 19 Tulsa Law Review, Vol. 14 [1978], Iss. 3, Art. 3

TULSA LAW JOURAAL [Vol. 14:491

show ways of looking at things, ways of sizing things up, ways of handling authorities, attitudes in one area of life-conflict and another."85 As a result, lawyers can ascertain the kind of attitudes likely to be manifested by the various members of the court and can be prepared to deal with these judicial value positions prior to arguing their cases. In other words, judges are restrained by their past behavior which manifests a particular set of values. Attorneys do not have to face a court without any idea of the range of attitudes and orientations they will encounter. The judge limits himself by the past legacy of attitudes he has manifested. This type of "attitudinal analysis" has been significantly under- taken by political scientists. The objective in this kind of study is both to understand the common values that act as a cohesive force holding blocs together and to become aware of the differences in values manifested within individual blocs. An important dimension for Llewellyn is thus the consistency of judicial attitudes. One example of this method of analysis is found in an article by Harold Spaeth. 6 His goal was to identify "ideational identity" between members of the Supreme Court from 1953 to 1958. The bloc used for analysis was that of Justices Black and Douglas. The analysis focuses on both the extent of voting agreement and the character of the issues which divided the usually highly cohesive pair. Through vote analysis Spaeth demonstrated certain variations between the pair even though they had a high overall index of agreement. For example, Douglas was more intense in his civil liberty feelings and more inclined to dissent than was Black. In suits between private taxpayers and the government, Black almost uniformly voted for the government, while Douglas supported the taxpayer. Black was more states'-rights oriented than was Douglas, while Douglas was more judicially active than was Black. In comparison with the more traditional case analysis, this vote counting method allows more accurate and discriminating comparison of a pair of Justices than could be attained by simply indicating the high levels of agreement between them. Another attitudinal study by Spaeth used more sophisticated techniques which go beyond the mere identification of a single important variable.87 Various studies-many of which have been discussed in

85. Id. 86. Spaeth, An Approach to the Study ofAttitudinalDifferences as an Aspect ofJudcial Beha - ior, 5 MIDWEST J. POL. SCI. 165 (1961). 87. Spaeth, JudicialPower as a Variable Motivating Supreme Court Behavior, 6 MIDWEST J. POL. Sci. 54 (1962).

https://digitalcommons.law.utulsa.edu/tlr/vol14/iss3/3 20 Clark: Karl Llewellyn on Legal Method: A Social Science Reconsideration

1979] KARL LLEWELLYN this essay-have identified crucial factors influencing judicial votes, such as attitudes about civil liberties, economic liberalism, federalism and, the court's supervisory capacity over lower courts. Spaeth investigated those case decisions in which such value-laden choices are not offered to see if a particular single factor can explain these remaining decisions. He suggested that views toward judicial activism/self-restraint could be such a variable, and he chose cases dealing with com- ity, civil procedure, diversity of citizenship, and venue as representing these value-free choices. Generally he found a strong correlation, and he explained apparent contradictions by using sub-category analysis to indicate that a particular Justice had-for varied reasons-a personally strong interest in a certain category of otherwise value-free choices. The next step beyond single variable attitudinal analysis has been taken by Glendon Schubert.88 He proposes a multitudinal model, com- bining into one framework a number of distinct values indentified by scaling. Instead of a hypothesis that argues judges act in civil liberties cases because of single variable, attitude toward civil liberties, Schubert in effect combines five or six scales (view toward judicial power; attitude toward the federal government, for example) into one three-dimensional ("five-dimensional factor space") model to delineate a distinctive configuration of attitudinal outlooks for a particular judge.8 9 Hence Schubert's model recognizes, and seeks to remedy, one basic fault of cumulative scaling-its uni-dimensionality. The judicial mind when faced with a civil liberties case does not automatically exclude the rest of the individual's experiences and values (a possible strong preference for judicial restraint, for example), though these attutides may play subordinate roles. Schubert's model recognizes this and reflects the interplay among and between a range of attitudes that occurs in an individual judge's mind. A further variation on attitudinal analysis is demonstrated in another article by the prolific Schubert.90 Rather that relating pairs and blocs of Justices to one another, this study utilizes recent methodological innovations to produce an extensive study of the judicial philosophy of a single justice. Schubert, in part, was seeking to disprove criticism

88. Schubert, The 1960 Term of the Supreme Court: A PsychologicalAnalysis,56 AM. POLITI- CAL ScI. REv. 90 (1962). 89. See also Schubert, Judicial 4ttitudes and Voting Behavior: The 1961 Term of the United States Supreme Court, 28 L.R. CONTEMP. PROB. 100, 119-32 (1963). 90. Schubert, Jackson's JudicialPhilosophy: An Explorationin Value Analysis, 59 AM. POLIT- ICAL SCL REV. 940 (1965).

Published by TU Law Digital Commons, 1978 21 Tulsa Law Review, Vol. 14 [1978], Iss. 3, Art. 3

TULSA LAW JOUR[1AL [Vol. 14:491

that such techniques cannot deal in detailed value analysis as the more traditional methods of judicial biography and case analysis. But Schu- bert argues that his techniques are better than the traditional methods because they manifest computer-precise data recall, permit simultane- ous study of all opinions, and result in consistent evaluation of all data. Using some esoteric methods, Schubert proceeds to test a number of hypotheses about Jackson's philosophy, proving some, disproving most. The methodology is based upon some thirteen content variables subject to a variety of mathematical operations. In his conclusion, Schubert is able to suggest some interesting distinctions between Jack- son's overall voting behavior as contrasted with his written opinions. It is clearly possible to determine the value orientation of judges toward a variety of issues by use of political science techniques. It is also obvious that there is some measure of persistence in these value orientations. These observations, however, only render Llewellyn's citation of value preference as a "steadying factor" more puzzling. The gist of his remarks seems to be that, while values do affect decisions, a properly prepared attorney who must confront appellate judges can ascertain these value preferences and prepare his arguments to meet them. Further, because these attitudes are persistent through time, the judge in effect has limited himself by his past behavior. The argument that this is a "restraint" overlooks several important points. First, the attorney will have to modify his presentation to comport with a particular court's legacy of past attitudes. Second, value performance is only a restraint as long as the judge chooses to adhere to particular attitudes, and he may very well shift his value orientation. Of all Llewellyn's steadying factors, this one most effectively indicates his dual approach to the issue of appellate decision-making. As a realist, Llewellyn con- ceived of the judicial process as being rent with discretion and flexibility. At the same time, he believed that there are some meaningful restraints (including the judge's own attitudes) which, while not as binding or constraining as logical determinism, still bring a measure of limitation to appellate decision-making. The issue remains, however, how much any given decision reflects value preferences rather than steadying factors. That the two are present does not excuse judges from ultimately simply transmuting their value preferences into judicial policy.

https://digitalcommons.law.utulsa.edu/tlr/vol14/iss3/3 22 Clark: Karl Llewellyn on Legal Method: A Social Science Reconsideration

1979] KARL LLEWELLYN

D. Some Other Steadying Factors To conclude this examination of Llewellyn's theses, a number of the remaining steadying factors will be lumped into one category. These points as identified by Llewellyn are (1) law-conditioned offi- cials9 (2) legal doctrine92 (3) known doctrinal techniques93 (4) issues limited, sharpened, phrased94 and (5) adversary argument by counsel.95 In total, these points seem to all relate to the fact that a judge is "social- ized" by his past training and experiences as a lawyer (and previous judicial service, if any) to deal with legal issues in a particular technical fashion. Socialization involves the setting of ones individual behavior patterns, and the appreciation of the values of institutional roles.96 Key steps in the judge's socialization process include law school, practice, interaction with others in the legal system, and bar association activities. Interestingly enough, the socialization process can continue after he assumes judicial office as well. For a federal judge, such factors as his colleagues on the bench, seminars,97 judicial council activity, sup- portive staff (particularly law clerks), and local lawyers, all tend to reinforce the new judge's technical socialization. These factors also relate to his role conception.98 As a result, the judge experiences constant reinforcement of his past training which emphasized a technical approach to legal issues. Certainly this is not the only factor that influences his decision-making, but undoubtedly years of past training and experience--constantly reinforced--exercise some influence upon a judge's handling of legal questions.

V. CONCLUSION Karl Llewellyn's goal in writing The Common Law Tradition: De- ciding Appeals was to restore some balance to the examination of appellate decision-making. Clearly, Llewellyn had not abandoned his realist orientation toward judicial policymaking, for to say that judges have discretion and flexibility is not to state that there are no restraints

91. LLEWELLYN, THE COMMON LAW TRADITION, supra note 5, at 18. 92. Id. at 20. 93. Id. at 21. 94. Id. at 29. 95. Id. 96. See generally R. DAWSON & K. BRawrrT, POLITICAL SOCIALIZATION (1969). 97. Cook, The Socialization of New FederalJudges: Impact on District Court Business, 1971 WASH. U.L.Q. 253, 263-66. 98. Carp & Wheeler, Sink or Swim" The Socializationof a FederalDistrict Judge, 21 J. PUB. LAW 359 (1972).

Published by TU Law Digital Commons, 1978 23 Tulsa Law Review, Vol. 14 [1978], Iss. 3, Art. 3

TULSA LAW JO URNAL [Vol. 14:491

upon their decision-making patterns. To maintain that legal logic does not absolutely control judges is not to assert that judges can freely ignore their training and technical orientations. The examples of political science literature discussed within this paper substantiate Llewellyn's intuition of the effects of judicial role and collegial decision-making upon appellate judges. The literature on attitudes also sustains his contention that it is possible to ascertain judicial attitudes and that these attitudes do tend to remain stable over time. And finally, the literature also supports Llewellyn's position that past legal skills and socialization tend to play a role as well. Generally, the modem social science approach manifested in judicial behavior studies confirms many of the early theses of legal realism on motivation of judicial decision-making. Although this paper has been able to discuss only a fraction of the available literature, it seems obvious that a substantial body of research findings now exists relating to judicial decision-making that can be drawn upon by any who are interested. The key problem which lies ahead is to develop more precise techniques so that it will be possible to evaluate the role which the kind of restraining factors Llewellyn suggested may play in any given decision or line of decisions. Llewellyn's intuitions about restraining factors have been substantiated by the subsequent ample body of judicial behavior research. It remains to be determined what weight any of these factors may play, vis-a-vis simple judicial preference, in decision- making. The weight of these factors may well vary, from judge to judge, and perhaps even relative to the same judge in different situations. So the trail blazed by the realists is being followed and developed, and the restoration of balance argued for by Llewellyn is also progressing. The resulting benefit is a more complete understanding of the judicial process.

https://digitalcommons.law.utulsa.edu/tlr/vol14/iss3/3 24